Options in "netVisual_embedding" and Summary of CellChat?

[phoebee-h]

Hi,
Thanks for such an amazing CellChat package! I wonder:

(1) Is there an option (eg. set.seed) for the UMAP plot of "Structural" or "Functional" network estimation, so that the results could be the same in each computation? (As below, even though they are mathematically the same...)

(2) Is there a chart for summarizing the results (eg. p-value or something for sorting), since I could get plenty of cell-cell communication possibilities?

Thank you,
Phoebe.

[sqjin]

@phoebee-h Thanks for bringing it out. I will look into this issue next week. For your second question, I am not quite understanding it. Can you check subsetCommunication function to see if it provides results you need?

[phoebee-h]

Hi,
I've checked the function subsetCommunication, which indeed helps! However, I'm not quite sure about the pval column:
(1) Why it only contains values under 0.05? But in the slot net, I can see far more than that?
(2) Why the p-values are not that continuous, which means they seem to categorize in certain intervals?

net <- subsetCommunication(cellchat, slot.name="net")

Thank you,
Phoebe.

[sqjin]

@phoebee-h Sorry, I missed your questions. (1) You can set thresh parameter to include more interactions; (2) We use a simple permutation test, where you can find it in our method section in the paper. This method will produce values like what you were showing.

[phoebee-h]

I see.
Thank you very much.